Most efficient numerical algorithms in scientific computing and machine learning require the user to provide a routine that can calculate the derivative of a quantity with respect to some underlying parameters. With the increasing complexity of the models used in mechanics and the rise of scientific machine learning (SciML), users are increasingly turning to automatic differentiation (AD) techniques [1, 2] that can automatically derive this routine. This minisymposium aims to bring together researchers, software developers and users of automatic differentiation techniques to discuss the state-of-the-art in the field of automatic differentiation and its application to various challenging problems in mechanics. Possible topics include: - the solution of challenging mechanics problems leveraging AD, - novel combinations of AD approaches, e.g. coupling PDE-level AD tools with source-level AD tools, - end-to-end differentiable numerical solvers and operator learning approaches including those using non-traditional basis functions, e.g. neural networks, - material constitutive modelling using AD, - checkpointing strategies for reverse-mode AD, - and AD software tools and frameworks. [1] Farrell, P., D. Ham, S. Funke, and M. Rognes. 2013. ‘Automated Derivation of the Adjoint of High-Level Transient Finite Element Programs’. SIAM Journal on Scientific Computing 35 (4): C369–93. https://doi.org/10.1137/120873558. [2] Naumann, U. 2011. The Art of Differentiating Computer Programs. Software, Environments and Tools. Society for Industrial and Applied Mathematics. https://doi.org/10.1137/1.9781611972078.